Post-Electoral Coalition Formation, Exclusion Risk, and the Winner's Threat: Theory and Simulation Evidence

Replication package for: "Post-Electoral Coalition Formation, Exclusion Risk, and the Winner's Threat: Theory and Simulation Evidence" Miguel Baigorri & Juan M. Benito-Ostolaza (Public University of Navarre, UPNA)Submitted to Public Choice (Springer Nature), 2026. This package contains all code, data, and figures necessary to replicate the results reported in the paper. The model develops a non-cooperative bargaining framework for post-electoral coalition formation and tests its predictions against a panel of 157 government formation episodes across 18 European democracies (1990–2023). Contents: code/simulations.py — Monte Carlo simulation engine, Tables 1–2, Figures 2–4, Appendix C code/empirical_analysis.py — Panel OLS with HC3 robust standard errors, Table 3, Figure 5 data/parlgov_full.csv — Empirical dataset (N=157, 18 countries, 1990–2023), derived from ParlGov (doi:10.7910/DVN/2VZ5ZC) data/ — All simulation draws and robustness checks figures/ — All manuscript figures (PDF) Requirements: Python 3.x, numpy, matplotlib, statsmodels, pandas, scipy To replicate: python simulations.py / python empirical_analysis.py